CHAPTER 4 EVALUATING STREET IMPROVEMENT
4.2 Route choice experiment
Figure 8 An example of IIA violation in a pedestrian’s route choice.
Referring to welfare measurement (Freeman, 1993), equation (17) estimates the change in WTW for the improved SA:
An additional issue is whether RP and SP data can be combined into a joint model. This can enlarge the sample size to assist insignificant variables to enrich data variation and improve the correlations between attributes in RP data. Swait and Louviere (1993) proposed a procedure for joint estimation through log-likelihood value testing, after rescaling the relative scale factor. This would be performed in the next step of this research.
4.2 Route choice experiment
Psychological measure is different from the physiological one (WEE). Walking time is more suited to a self-reporting survey, because respondents can recognize time spent, but not distance traveled. Thus, I denote walking time as the measure of WTW. A higher WTW is
36
considered as an effect of improved amenities from improvement projects. To systematically describe alternative routes, each attribute of SA is represented with two or three levels, as shown in Table 9.
Table 9 Attributes and levels of SA.
Attribute Level Description
The degree of lighting that street lamps offer pedestrians to ensure safety after dark.
Planting(PT) Potted plants(P1) Bushes(P2) Trees(P3)
Three types of planting on street (reflecting green level).
Street
Furniture(SF)
Few(FF) Various(FV)
To indicate whether a street offers various facilities to serve more outdoor activities for pedestrians.
Pavement(PV) Asphalt/concrete(A) Colored brick(C)
Two types of walkway surface; the first is rigid, monolithic, and poor-colored; the second is flexible and colorful.
Retailing(RE) Few(RF) Various(RV)
To indicate whether the street offers retail services, such as food stands, kiosks, or a grocery stop.
Fountain(WA) None(NO) Available(AV)
To indicate if there is a static or flowing waterscape on street.
a. Levels of each attribute are coded using effects codes. The base level is assigned -1 to represent the lowest level of amenity. For 2-level attributes, the values are assigned -1 and 1; for 3-level attributes, the values are assigned -1, 0, and 1.
37
Definitions of levels and questions was revised by a 14-person focus group consisting of four students, six researchers, and four residents, interested in our study, to ensure that questions are designed to give participants clear information consistent with what the study wants to ask. It is also important for them to make questions to be user-friendly to help the participants form judgments. For example, right of way for pedestrians is a major determinate of street design (for example, in sidewalks or arcades) to exclude vehicle or incompatible use; after revision, it refers to whether the street protects pedestrians from a conflict between “exclusive” or “shared” use. Lighting provides safety for pedestrians after dark, and there are three Lux levels: 5 Lux, 10 Lux, and 15 Lux, to represent the degree of lighting here. The common types of road planting - potted plants, shrubs, and trees - are used to represent the degree of greenery and enclosure. Street furniture is a set of equipment with various functions for pedestrians, but only includes benches, signs, waste receptacles, and public art, in this instance. The pavement is made from two common types of material:
asphalt/concrete and bricks. Street retail usually consists of food stands, kiosks, sidewalk cafés, retail stores, and so forth. In this case, its levels are designated as “few” and “various”
to indicate whether the street accommodates a variety of daily need. Lastly, the design of a street fountain will feature flowing water or a static pool; here, we use “none” and
“available” to indicate whether there is static or running water on the street.
Five variables - gender, age, weight, income, and education - are considered as individual characteristics. The levels of services (A~F) defined in Fruin (1971) are used to represent the effects of free-flow degree on pedestrians’ utility (see Appendix 1). Stay home or drive vehicle (HO) is the alternative specific constant in the SP questions, used to indicate whether respondents accepted the alternatives designed for walking.
The RP question is designed to record people’s actual route choice behavior. However, a pedestrian chooses one potential walking route from the given O-D pair near his/her origin.
If the street conditions of those routes are similar, there would be little variation in the attributes of SA between alternatives. This would prove a disadvantage in our model estimation, since a discrete choice model is operated by detecting the degree of variation between each attribute (Louviere, Hensher, and Swait, 2000). Otherwise, RP data may suffer from colinearity among the attributes of alternative routes (Adamowicz, Louviere, and Williams, 1994). To solve this, the SP question is designed to elicit preferences by using hypothetical questions, whereby alternatives and choices would not be restricted in the status (Louviere et al., 2000). For example, street furniture is rarely placed on a shared
38
ROW walkway in Taiwan, but the SP question could simulate this situation. In addition, earlier studies found that elicited preferences from multiple sources would predict behavior better than single ones (Huber, Wittink, Fiedler and Miller, 1993; Johnson et al., 2000;
McFadden, 1986).
Table 10 List of the individual characteristics and the specific constant.
Characteristic Level Description
Gender(GE) Male(GE=1)
Female(GE=0)
Dummy variable to indicate difference between genders
Age Years Number of years that a person has lived
Weight(WE) kg Total body weight
Income(IC) NT$/per month Gross monthly income Education(ED) Elementary
Level of Service(LOS) Level A(LOS=1) Level B(LOS=2)
…
Level F(LOS=6)
Degree of free walking according to Fruin (1971) definition
Stay home or drive vehicle(HO)
Yes(HO=1) No(HO=0)
A specific constant for each SP choice set to detect whether respondent accepts the two designed alternatives
The survey trips include work-to-home, home-to-shopping, and home-to-leisure; unlike home-to-work trips, people have more free time and discretion in those trips to decide between walking time (distance) and SA, thus they will not necessarily choose the shortest path. In the first part of the questionnaire, respondents are asked to give checks on the individual characteristics in Table 10 as a warm up task. They then describe their trips by answering RP questions regarding origin, destination, most frequent route, two alternative routes, and road names. A map is also prepared for respondents for when they could not accurately recognize road names or alternative routes. The attributes and levels in Table 9
39
was designed as a checklist for respondents to describe the SA of each route. If the route involves several links of different street space, the links are described respectively, and the value of xink is calculated by taking the average of links of the various street types weighted by link length.
The SP question was designed using the fractional factorial method (Louviere et al., 2000).
Instead of using full-factorial design to produce a total of 2533 alternatives, this study applies SAS 9.0 software to design the five 2-level and three 3-level attributes into a balanced and orthogonal matrix. As a result, 36 alternatives are generated with D-efficiency 100% (see Table 11), meaning the colinearity was eliminated (Kuhfeld, Tobias, and Garrat, 1994). Note that walking time is regarded as an attribute when designing alternative, because it represent the cost of each route and people should take it into account when choosing the route. It is categorized into three levels: 5 (convenience), 10 (acceptable) and 15(long) min with reference to literature (Pushkarev and Zupan, 1975).
Figure 9 An example of a stated preference question
40
Table 11 List of the designed SP alternatives.
Block Set Alternative ROW LT PL SF PV RE WA WT
1 1 1 1 3 2 2 1 1 2 3
1 1 2 2 1 1 1 2 2 1 2
1 2 1 1 1 1 2 2 2 2 1
1 2 2 2 2 3 1 1 2 1 2
1 3 1 2 2 3 2 2 2 1 1
1 3 2 1 1 2 1 1 1 2 2
1 4 1 2 3 2 1 1 1 2 1
1 4 2 1 2 1 2 2 2 2 3
1 5 1 2 2 3 2 2 1 2 2
1 5 2 1 3 1 1 1 2 1 3
1 6 1 1 1 3 2 1 1 1 3
1 6 2 2 2 2 1 2 2 2 1
2 1 1 1 1 3 2 1 2 2 1
2 1 2 2 2 1 1 2 1 1 3
2 2 1 2 2 2 2 1 2 2 3
2 2 2 1 3 1 1 2 1 1 1
2 3 1 1 2 3 1 2 1 2 2
2 3 2 2 3 2 2 1 2 1 1
2 4 1 2 1 1 2 1 1 1 2
2 4 2 1 2 2 1 2 2 2 1
2 5 1 1 3 3 1 2 1 2 3
2 5 2 1 2 1 2 1 2 1 2
2 6 1 2 3 1 2 2 1 2 2
2 6 2 1 1 3 1 1 2 2 3
3 1 1 2 3 3 2 2 2 2 2
3 1 2 1 2 1 1 1 1 1 3
3 2 1 2 1 2 2 2 1 1 3
3 2 2 1 3 3 1 1 2 2 2
3 3 1 2 1 1 1 1 1 2 1
3 3 2 1 3 2 2 2 2 1 2
3 4 1 1 2 1 2 1 1 2 2
3 4 2 2 1 3 1 2 2 1 3
3 5 1 2 2 3 2 1 1 2 1
3 5 2 1 1 2 1 2 2 2 2
3 6 1 1 2 2 1 1 1 1 1
3 6 2 2 3 1 2 1 2 2 3
41
Those alternatives were further blocked into three groups (12 alternatives in each group), so each respondent only need to make six choices (two alternatives in a choice set). Two techniques are used to aid respondents when making judgments; the initial condition with zero travel cost could be used as the reference of other alternatives, and the figure of street space could help respondents to understand what the street looks like, as shown in Figure 9.
However, the alternative would violate IIA because it implies that respondents decide not to walk, and it is truly different from the other two alternatives. This is another reason why I used RPL. Finally, the total number of questions were reduced to less than 21, and the SP choice sets were less than 8, as suggested in empirical studies (Kuhfeld, 2005).